Ground-state chromophore geometry, not cage size, tracks quantum yield in fluorescent proteins

A common explanation for brightness variation in fluorescent proteins (FPs) is that tight chromophore cages produce bright proteins. This model predicts that quantum yield should track the dihedral rotational space available to the chromophore. We tested that prediction across 838 FP crystal structures from the Protein Data Bank. For each structure, we scanned the chromophore's two methine-bridge torsions, τ and φ. We measured two features of the chromophore environment: the fraction of torsional space that is sterically accessible, and the angular distance from the deposited chromophore to the nearest planar geometry. The scan shows a consistent family-wide asymmetry. The I-bond, the cis-trans isomerization axis, is clamped in all color classes. The P-bond, the ring-rotation axis, is the variable degree of freedom. Gatekeeper analysis identifies position 203, in avGFP numbering, as the dominant P-bond constraint. It is the first steric barrier in 21% of all sweep events. Cage size does not generally track quantum yield; instead, it primarily reflects chromophore chemistry. Indole-based cyan FPs and acylimine-based red FPs have the tightest cages, while imidazole-based blue FPs have the loosest. In red FPs, quantum yield is associated instead with ground-state chromophore geometry. More twisted crystallographic chromophores are consistently dimmer in red FPs. The scan also shows that rigidity is not a single property. Cage size, ground-state geometry, gatekeeper residues, and thermal immobilization are distinct measures of rigidity, each with a different relationship to quantum yield. Fluorescent-protein design should therefore target local control of chromophore geometry rather than global cage tightening.